POLICE STATISTICS RECENT DEVELOPMENTS IN FEDERAL CRIMINAL STATISTICS. Dana M. Barbour, Office of Statistical Standards, U. S. Bureau of the Budget

Transcription

1 42 RECENT DEVELOPMENTS IN FEDERAL CRIMINAL STATISTICS Dana M. Barbour, Office of Statistical Standards, U. S. Bureau of the Budget Several improvements have been made during the last few years in the criminal statistics- - police, judicial and correctional -- issued by the Federal government. In this paper I propose to describe briefly the existing Federal series in the field, both those based on data collected from State or local authorities and those relating to Federal offenses, paying particular attention recent developments and noting some plans for further improvement. I want to point out certain limitations which are not always understood by some of the users of these data. POLICE STATISTICS In the area of police statistics we have the Uniform Crime Reports, which attract the most attention and are probably the most widely used of any of the Federal series on crime and correction. While the reporta are published by the Federal Bureau of Investigation, the program is a cooperative one with the International Association of Chiefs of Police. Begun 30 years ago the number of reporters has increased over the years until about 7,000 State and local police departments, sheriffs' offices and other police agencies are now sending monthly and annual reports to the F.B.I. Recommendations of Consultant Committee on Uniform Crime Reporting Late in 1957 the F.B.I. appointed a Consultant Committee on Uniform Crime Reporting to make an independent analysis of the program. The committee was headed by Dr. Peter Lejins of the University of Maryland. The other members were Dr. Charlton F. Chute, Director of the Institute of Public Administration, New York City, and Col. Stanley R. Shrotel, Chief of Police of Cincinnati. Their report was published in 1958 as a special issue of the Uniform Crime Reports. It contained 22 recommendations, all of which have been accepted at least as ultimate goals by the and the F.B.I. Only the major ones can be considered here. One of the most important recommendations related to expansion of coverage of the statistics on offenses known through arrests, which did not include rural areas at all, and which had represented only about percent of the urban population. the 1958 report was issued, Mr. Jerry Daunt, the new chief of the Uniform Crime Reporting Section of the F.B.I., reports some progress, though he thinks the data for rural areas are not yet representative enough to justify publication in the 1959 annual report. The Consultant Committee considered sampling as an alternative to attempting to collect information on crime for the entire country. While not ruling out sampling as a possibility the F.B.I. is not able to build up its reporting, the committee inclined to favor complete coverage if that could be attained. The reasons were the interest of police and the public in crime figures for individual communities and States and the almost complete coverage already achieved for a part of the Uniform Crime Reports- -the offenses known to police. Another important change resulting from the work of the Consultant Committee is the use of annual estimates of population instead of the last decennial census figures in computing crime ratea. Because of the differential population growth in the United States, the continued use of the 1950 figures had made the crime rates in States like California and Arizona look worse than they were, while having the opposite effect in States or cities with stable or declining populations. A somewhat similar change was the replaceient of the old urban -rural classification used in p resenting statistics with three -way classification of standard metropolitan areas (now standard metropolitan statistical areas), other cities, and rural areas. An important change in publication policy was the shift from semiannual bulletins to an annual bulletin. This recommendation was made to permit more time for verification of data and statistical analysis, and to eliminate a certain amount of overlap between the semiannual reports. The first annual bulletin under the new plan appeared in September 1959 for the calendar year These annual bulletins are being supplemented by a four -page quarterly release containing percentage changes in the city crime index and preliminary figures by offense for cities over 100,000. Finally the so- called Class I offenses used in measuring crime trends have been reexamined and some changes made as the result of recommendations of the Consultant Committee. The offenses now included are murder and nonnegligent manslaughter, forcible rape, robbery, aggravated assault, burglary, larceny --$50 and over in value, and auto theft. Negligent manslaughter was dropped partly on the ground that since it was made up almost entirely of culpable but nevertheless accidental traffic deaths it does not seem a proper offense to be used in an index of crime, and, pertly, because the police are frequently hesitant in classifying an accidental death as manslaughter and tend to wait for action by the grand jury. Larceny under $50 was also dropped for a number of reasons, the most important being that it is often a minor offense which is not reported to the police. Beginning with 1958 the F.H.I. on its own initiative excluded statutory rape from its Class I offenses on the ground that this was an offense which was seldom reported to the police. The seven offenses remaining are included because they meet several criteria reasonably well. They are serious offenses; they are usually reported to the police (though even among these seven offenses the completeness of reporting varies); their definition is relatively uniform throughout the country; and they are sufficiently numerous to be

2 43 statistically significant. They include offenses against both person and property, and they also are of importance to the cooperating police departments. Uniform Crime Reports as a Measurement of Crime We may ask how well this revised list of seven offenses measures trends in crime. More specifically,is the F.B.I. justified in stating, as it did in its last annual report, that "crime in the United States showed an over -all increase in 1958 of 9.3 percent over the 1957 level," or even, as the most recent quarterly release does, that preliminary figures for the city crime index in the first half of 1960 show an increase of 9 percent over the corresponding half of 1959? My own opinion is that while the trends in the United States for each of the seven offenses shown are fairly reliable, we are still far short of valid index of crime as a whole in the United States. Perhaps such an index can never be compiled. Implicit in the F.B.I. index is the assumption that the seven offenses measure, not merely all types of offenses known to the police, but all crime -- reported or unreported --in the United States. While the seven index offenses are certainly important ones, such serious types of offenses as forgery and counterfeiting, embezzlement, drug peddling and drunken driving are not included. No direct numerical comparison between the total number of offenses known to the police and the total shown for the seven index offenses is possible because (with minor exceptions) the F.B.I. does not collect data on offenses for other than the seven. If, however, we use the figures on arrests by offense (which are less satisfactory because they are less comprehensive), we find that of the total of 2,340,000 arrests in 1958 only 256,000 were for the seven offenses shown in the index of crime. If we subtract from this 2,340,000 total arrests on suspicion, for vagrancy, for disorderly conduct, and for drunkenness (on the ground that these are either not crimps at all or are not very serious ones), we still have almost three times as many arrests as the total shown for the index offenses. Moreover, several of the offenses not included in the index are covert rather than overt offenses, and as such are less likely to become known to the police. This is particularly true of embezzlement, fraud and sex offenses. We most conclude that in attempting to measure crime we are dealing with an unknown universe, of which the offenses known to police represent only part. I have not seen any evidence to demonstrate what, if any, correlation of movement exists between these seven offenses and total crime in the United States. As Professor Griffin put it four years ago, "The relationship between known crime and total crime is unknown, and is, in part at least, a function of the intensity of police activity. One more point maybe noted before leaving this matter of a crime index. This is the wide range of criminal behavior that is covered by the seven offenses included in the F.B.I. index, and the lack of any weighting by seriousness of the offense. For index purposes a $50 larceny is equated with a premeditated murder. Use of the Uniform Crime Reports for Interarea Comparisons Wide use is also made of the Uniform Crime Reports for interstate or intercity comparisons. For the seven index offenses this is facilitated by the publication in the annual bulletin of the rate per 100,000 inhabitants for each offense and all seven combined by States and by standard metropolitan statistical areas. The dangers of interarea comparisons are recognized by the F.B.I. itself. In the last annual report this statement is underlined: "Caution should be used in comparing crime data for individual cities because the differences in the figures maybe due to a number of factors. Such comparisons are not necessarily significant even though the figures for individual communities are converted into terms of offenses per inhabitants. " The report goes on to say t it is more important to determine whether the figures for a given community show increases or decreases than to ascertain whether they exceed or fall short of those for some other individual community. The Chairman of this session, Ronald Beattie, Chief of the California Bureau of Criminal Statistics, has devoted considerable attention to use of the Uniform Crime Reports for interarea comparisons in an article in a recent issue of the Journal of Criminal Law, Criminology and Police Science He notes that the 195d report shows has the highest crime rate of any State in the country and Los Angeles the highest of any metropolitan area, and suggests two possible explanations: one is California's broad definition of burglary, which includes acts that many other States would classify as larceny under $50; the other is the high standard of police efficiency in the State, the Los Angeles Police Department in particular being one of the best in the country. These high standards include good recording and reporting systems. Ronald Beattie's own organization, the California Bureau of Criminal Statistics, does a good deal to promote high standards John I. Griffin, "New Perspectives in Police tatistics," Journal of Criminal Law, Criminology and Police Science, Vol. 46, March April 1956, p "Uniform Crime Reports," 1958, p. 79. Ronald H. Beattie, "Criminal Statistics in he United States ," Journal of Criminal Law, Criminology and Police Science, Vol. 51, May-Jane 1960, p. 49.

3 44 of recording and reporting among the California police departments. Of course, the highly urban character of California's population is also a factor. In his article Beattie points out figures and ratios for other States than California and cities other than Los Angeles that appear suspect. To cite an extreme example there were only five cases of aggravated assault in Vermont in 1958 for a rate of 1.3 per hundred thousand inhabitants. There were more than twice as "many murders and nonnegligent manslaughter cases, the rate being 3.2. The national rates, on the other hand, are 65.5 for aggravated assaults and 4.7 for murder and nonnegligent manslaughter. This suggests that police authorities in Vermont are just not classifying assaults_ properly in their reports. fairness it should be said that the F.H.I. has devoted a great deal of effort to obtaining reports from police departments according to its uniform classification of offenses. A "Uniform Crime Reporting Handbook" is given to participating police agencies, and special instruction bulletins on particular subjects are issued from time to time. F.B.I. agents are available to visit local police departments where needed to help in the preparation of reports. There have been instances where the F.B.I. has omitted data for a city with the footnote, "The crime reporting for the city indicated does not meet acceptable standards established by the Committee on Uniform Crime Records of the International Association of Chiefs of Police." (Incidentally, it may be noted that such a footnote has proven a powerful stimulus in improving the reports of the city in question.) Despite these efforts to maintain high standards, it is true that, as the F.B.I. itself puts it, "The basic responsibility for the accuracy of the reports rests with the local law enforcement executives from whom and for whom the data are collected. "W Particular caution must be used in interpreting the data on juvenile arrests in the tables on arrests by age and offense. The practices of jurisdictions differ widely in their treatment of arrests of juveniles. Some will not report detention of a juvenile as an arrest at all. This makes impossible interarea comparisons. As regards trends, the increasing emphasis on juvenile delinquency and the creation of special units in police departments to deal with juveniles may account in part for the increase in the number of arrests. Statistics on Federal Crime We should not leave the area of police statistics without discussing briefly statistics on Federal crimes. These may be divided into two classes: ordinary offenses which become Federal crimes because they are committed on Federal property or involve interstate commerce, and specialized offenses which are peculiarly Federal because they involve violations of Federal laws "Uniform Crime Reports," 1958, P. 21. such as the immigration, internal revenue, postal, customs, and pure food and drug laws. Probably most of the ordinary offenses are included in the F.B.I. Uniform Crias Reports, without being identified as Federal offenses. The theft of an automobile will be reported as such by the local police department without waiting to determine whether a Federal crime may also have been committed because the automobile bed been driven across the State line. The F.B.I. also makes an effort to include offenses committed on Federal reservations. However, specialized Federal offenses are not included in the Uniform Crime Reports. Usually such offenses are not reported anywhere until an arrest has been mode, which may be done by officers of the Federal agency having the responsibility for enforcing the particular law. some of these agencies, such as the Immigration and Naturalization Service, do publish data on arrests, ordinarily a peculiarly Federal offense does not become a "statistic" until it is reported by the U. S. Attorney's office. JUDICIAL STATISTICS Caning now to the field of judicial or court statistics, we find a large gap. There are at present no national data on the operations of State and municipal courts in processing criminel cases involving adults. Census Series on State Judicial Criminal Statistics From 1932 to there was a series on State judicial criminal statistics compiled by the Bureau of the Census. The coverage of this series vas never truly national. In its last year only 25 States were cooperating, and not all counties were included even in those States. In addition to incomplete coverage, there was lack of comparability in the data because of variations among States with respect to the types of courts handling felony cases and because of differences in the offense classifications used. This last true despite the efforts of the Census Bureau to obtain uniform reporting according to a standard classification. Some of the data were of questionable reliability, largely because most clerks of court, who prepared the reports initially, insisted on using a tally -sheet method of reporting cases. There were long delays in publishing some of the reports. Finally, the Census Bureau appears not to have attached much importance to judicial criminal statistics and to have devoted very lit tle staff or attention to them. In 1946 the series was discontinued by the Census Bureau, after consultation with a committee of experts convened by the Office of Statistical Standards. The experience does illustrate the great difficulties faced by any Federal agency which tries to compile data based on reports from thousands of jurisdictions with varying types of organizations, laws and recordkeeping practices. Juvenile Court Statistics National juvenile court statistics have been collected and published by the Children's Bureau for more than 30 years. It was not until 1956, however, that it could be said with confidence

4 4.5 that the number of cases reported was representative of the country as a whole. Some regions of the country had been underrepresented, others overrepresented, because the statistics depended on the cooperation of the juvenile courts. The first report based on a national sample of courts was for the year This sample utilized the Census selection of 230 primary sampling units, each consisting of a county or several contiguous counties, for its own Current Population Survey. The selection was made on a random basis, after stratifying the primary sampling units on the basis of geographical location, population density, rate of growth, racial composition, economic characteristics, etc. It was found that there were 502 courts in these 230 primary sampling units. Estimates of the number and type of cases for the United States are made, based on the reports from the sample courts. The rate of delinquency cases per 1,000 children over age 10 is also computed by type of court (urban, semiurban and rural). These data are supplemented by reports from a number of courts not included in the sample. The cooperation of the courts included in the national sample has been good, but occasionally there are delays in submitting the reports. The policy of the Children's Bureau has been to work through the State agencies (State Welfare Departments, Youth Authorities or Attorneys - General). While this avoids possible duplicate reporting and strengthens the State agency, it has sometimes been a cause of delays in publication. Unfortunately, the data obtained from the national sample of courts are meager, being limited to the number of cases disposed of and the method of handling the cases -- official or unofficial. Traffic cases have been reported separately since No information is obtained on the reason for referral or the method of disposition of the cases, nor is there anything on the characteristics of the juveniles brought before the courts. When the national sample of courts was established it was deemed advisable to keep reports to a minimum in order to secure cooperation. The Children's Bureau hopes later to expand the information called for, but has no definite plans at present. It should be noted that last year the Children's Bureau and the National Institute of Mental Health of the Public Health Service made a special study of juvenile delinquency at the request of a subcommittee of the Committee on Appropriations of the House of Representatives. This study involved a number of one -time statistical surveys calling for much more detail than is available in the regular series. The results of these special surveys have been summarized in published hearings of the House Committee W and are also appearing as a series of pamphlets on juvenile "Report on Juvenile Delinquency," Hearings efore the Subcommittee of the Committee on Appropriations, House of Representatives, 86th Cong., 2nd.Sess. (1960), pp delinquency issued by the Children's Bureau./ Federal Court Statistics When it cames to judicial statistics for Federal courts we are in a much better position. The reports of the Department of Justice go back to Shortly after its creation in 1939, the Administrative Office of the U. S. Courts took over judicial statistics. These statistics are based on a separate card for each case prepared by the clerks of all Federal district and circuit courts. The information is put on punch cards, and quarterly and annual reports are prepared. The annual report of the Director of the Administrative Office of the U. S. Courts shows by circuit and district the number of criminal cases commenced, terminated and pending at the close of the year, and the number of defendants involved; information on cases and defendants by nature of offense; and criminal defendants disposed of, by nature of offense and disposition, including length of sentence for those sent to prison. Separate figures are presented for juveniles. In addition there are a number of tables on probationers under the supervision of United States probation officers, showing the number received for supervision and removed from supervision during the year by districts; the caseload of probation officers; the age, race, sex and offense of probationers; and the number of parolees as well as probationers who were reported as violators during the year. The Annual Report of the Attorney General also includes some judicial statistics, as well as detailed tables showing by judicial district the number of criminal and civil cases handled by United States Attorneys and their status at the end of the year. These statistics are based on reports from the United States Attorneys. CORRECTIONAL STATISTICS The third and last field is that of correctional statistics. National Prisoner Statistics The National Prisoner Statistics program has been discussed by Mr. McCafferty of the Bureau of Prisons,/ so I shall not go into it here. I would like to emphasize two encouraging developments- -the progress being made in reducing the backlog and putting the detailed statistics on court commitments and prisoners released from State and Federal institutions on a more current basis,and the use by the Bureau of Prisons of 6 "Juvenile Delinquency: Facts, Facets," o., (subtitle -- subject of report), U. S. Del:WA:lent of Health, Education, and Welfare, Social Security Administration -- Children's Bureau. / "Prisoner Statistics -- National and State," Paper read at the Statistics of Crime and Correction session, Annual Meeting, American Statistical Association, August 23, 1960.

5 46 punch cards which are being prepared by an increasing number of the statistical units of State central correctional agencies. Prisoners in Jails and Local Institutions The national prisoner Statistics program does not include prisoners in jails and other local correctional institutions. The only coverage in this area is that of the Census Bureau, which enumerates prisoners along with the rest of the population in its decennial census, and publishes statistics on them in its special report on institutional population. The report for the census was published in 1953,Y and presented considerable data on the age, sex, colo, etc: of persons in local jails and workhouses by State. No information was obtained on offenses or whether inmates were waiting trial or serving sentences. The Census Bureau plans to present about the same type of information for the 1960 census. However, the report on institutional population will probably not be published until late 1962 or early Juvenile Delinquents Mention should be made of the statistics which the Children's Bureau compiles on public training schools for juvenile delinquents.2/ These statistics were put on a regular annual basis in 1956 and include data on movement into and out of the institution, the status of the institutional population at the end of the period, and some information on staff and expenditures of these institutions. Certain supplementary data on training schools have also been obtained from time to tine. Federal Prisoners Statistics on Federal urisoners are included in the National PrisoflÑ Statistics program. In addition the Bureau of Prisons publishes considerable information about them in its annual report, "Federal Prisons." This includes data on characteristics of the prisoners, length of sentence and average time served by those released by type of offense. One important type of information not now obtained for prisoners in State institutions is gotten for Federal prisoners. This is information on recidivism by type of offense for prisoners committed to Federal institutions during the year. Statistics on parole of Federal prisoners are included both in the "Federal Prisons" report and the report of the chairman of the Board of Parole contained in the Annual Report of the Attorney General. These show parole decisions and parole grants by offense, average time served prior to parole, and violations by type of offense. PIANS FAR FURTHER IMPROVEMENT Thus far in this paper we have discussed what Federal criminal statistics are presently available, noting particularly recent developments. Brief mention might be made of a plan for better -integrated statistics on Federal crime developed by Ronald Beattie when he was in Washington last winter. The system would begin with complaints filed with the U. S. attorneys, and involves getting some additional information about' complaints not prosecuted and more detail about defendants. It also involves agreement between the Department of Justice and the Administrative Office of the U. S. Courts on uniform classification of data and uniform presentation of such data. Beattie points out that in almost every Federal case the offender is fingerprinted and given an F.H.I. number. If this number could be put on all documents relating to the offender throughout the criminal process, the problem of identification would be solved and it would be possible to show for X number of alleged Federal offenders, the proportion not prosecuted, the proportion convicted by the trial court, the proportion receiving fines or placed on probation, the proportion sent to prison, etc. It should be emphasized, however, that this plan is still in the proposal stage. My own office, the Office of Statistical Standards in the Bureau of the Budget, is planning to take the lead in creating an interagency committee of Federal statisticians in the criminal statistics field. We have noted the need for closer coordination in dealing with a number of problems. One of the problems such a committee might consider is the development of a single standard classification of Federal offenses. CONCLUSIONS In conclusion, I have tried to indicate that, despite their limitations, there have been a number of encouraging developments during the last few years in the criminal statistics compiled and published by the Federal government, and that there are plans for further improvement. It may well be, however, that filling the major existing gaps in this field must await further progress by the States. As Thorsten Sellin put it ten years ago: "... It is more than likely that we can never hope for a further fundamental improvement in the structure of a system national statistics based on voluntary cooperation until we have strengthened the foundation. That foundation must be laid in the individual states."12/ "Institutional Population," United States usus of Population: 1950, Special Report P -E No. 2C. 9/. Published under the title, "Statistics on Pub - Institutions for Delinquent Children, (year)," Children's Bureau Statistical Series, No.. Thorsten Sellin, "The Uniform Criminal tatistics Act," Journal of Criminal Law and Criminology, Vol. ho, March April 1950, p. 683.